196 ALASKAN GLACIER STUDIES
test of actual climatic records. It would certainly seem remarkable, though, if after a long period during which the precipitation was not sufficient to push the glaciers out as far as their present termini, there came a sufficient increase in precipitation to push glacier fronts forward 15 or 20 miles or more, and then after a brief interval, sufficient diminution in precipitationj or increase of temperature, to allow the glacier fronts to recede 15 or 20 miles. Such a great climatic change would of itself be so remarkable that we ought to at least see a sufficient reason for it. Failing that, the hypothesis seems improbable, and the improbability is increased by the discovery of the fact that the change in the glacier fronts of Yakutat Bay, and those north of the St. Elias Range and in Glacier Bay, have not been paralleled a few hundred miles to the northwest, in Prince William Sound, which lies in the same general climatic belt and whose heavy precipitation is induced by the same meteorological conditions that cause the heavy precipitation in the Yakutat and Glacier Bay regions. While we do not feel that in this case, as in the case of the hypothesis of uplift, we can dismiss the hypothesis as inadequate, we nevertheless are of the opinion that it is highly improbable and to be resorted to only when no other more adequate explanation is possible.
The third hypothesis is that of earthquake shaking, which at first thought, might seem both extravagant and impossible, but in favor of which, nevertheless, much can be said. By this hypothesis it is assumed that a century or more ago there was a great and long-continued shaking of the snow-covered ranges of the mountain region from Yakutat Bay to Glacier Bay, perhaps in a series of years, with repeated great earthquake shocks. By t.Tiia shaking and avalanche overloading of the glaciers a wave of advance was started, and in consequence of each great shock the advance was continued, until the glaciers reached their outermost positions; then, with cessation of vigorous shaking, the glaciers receded. The earthquake record unfortunately does not go back quite far enough for correlation as to this possibility, the earliest recorded earthquake which we have seen even mentioned having occurred in 1788.
This hypothesis has the one great merit that the other hypotheses lack: it would account for all the phenomena. By it we can explain the local nature of the advance, its absence to the northwest, and even the present advanced position of Columbia Glacier. The hypothesis would also account for the briefness of the advance, and for the subsequent rapid recession. There appears to be but one important objection to the earthquake hypothesis, and that is the stupendous response to it by the glaciers. It may be said, however, that the same objection applies also to all other hypotheses. There certainly was a great advance, difficult to explain by any of the hypotheses, and perhaps no more difficult under this one than under the others. In 1906 the senior author was inclined to believe this hypothesis so improbable that he resorted to "the vague hypothesis of climatic variations," but a further consideration of the problem, and especially the remarkable advance of Hidden Glacier, whose front was pushed 2 miles farther out in 1909 than it was in 1906, has led both of us to the belief that the earthquake hypothesis is not so utterly improbable as to warrant discarding it. It is certainly conceivable that a series of severe shakings, like those of 1899, repeated through successive years, might cause even so great an advance as the 15 to 20 mile advance of Hidden and Nunatak Glaciers. When we know the nature of the response that the Hubbard and Nunatak Glaciers are going to make to the earthquake shaking of 1899 we may be better able to consider this hypothesis. If they do not advance greatly we should consider it far less